Helpful post- diagnostic services for young onset dementia: Findings and recommendations from the Angela project

There is a significant lack of evidence regarding optimum models for service provision in young onset dementia (YOD). Our study aim was to gather detailed information about services experienced as helpful by those with YOD and family carers. The objective was to identify the core features of these services to inform service design, delivery and improvements. A qualitative approach based on appreciative inquiry was used, posing open- ended questions about services experienced as helpful, as part of a national UK survey of people with YOD and carers. We used inductive thematic analysis to analyse the free- text responses. The resulting template was used as a basis for analysis of in- depth follow- up interviews,

a. find person-centred interaction helpful: when staff have a positive attitude, are collaborative and flexible, and provide in-person support and user-friendly information.
b. find services helpful when they take younger age into account, consider the whole family, are accessible and respond proactively to changing needs. c. find service systems helpful when there is specialist young onset dementia provision, continuity over time and integration across services.

| INTRODUC TI ON
Young onset dementia (YOD) refers to dementia with onset under 65 years (Draper & Withall, 2016). Approximately 42,500 people live with YOD in the United Kingdom (Prince et al., 2014), accounting for 5%-9% of dementia prevalence. YOD is significantly different from dementia with late onset (LOD), due to the earlier stage of life and higher prevalence of rarer types (Vieira et al., 2013), resulting in distinctive needs. Those with YOD need active ways to continue involvement and contribution to life; support to manage difficulties associated with vision, communication, social cognition and movement; advice on employment and finances; and support for families with young people (Chaplin & Davidson, 2016;Mayrhofer et al., 2018;Millenaar et al., 2016;Roach et al., 2016;. Despite the distinctive needs, specifically commissioned YOD services are few, and short-term commissioning means many are later discontinued (Cations et al., 2017;Mayrhofer et al., 2018). Most people with YOD who receive post-diagnostic support, access this through all-age dementia services , with the exception, in the United Kingdom, for dementia caused by Down's syndrome, traumatic brain injury, HIV, Huntington's disease or alcohol-related dementia, whose services are usually provided through other specialist services. People with YOD are not well served by most 'all-age' services, which are geared towards older people with amnestic symptoms, resulting in lack of support for those with rarer dementias and provision of therapeutic activities that do not suit younger generations (Cations et al., 2017;Giebel, Eastham, et al., 2020;Mayrhofer et al., 2018;Sansoni et al., 2016;. Sansoni et al. (2016) report a catalogue of shortcomings: lack of a care pathway, poor provision of information, unmet needs for social companionship, day-time activities and lack of help regarding intimate relations, behavioural issues and communication.
There is little evidence regarding optimum service provision (Carter et al., 2018). In Australia, Westera and Fildes's (2014) literature review and stakeholder consultation suggested services should be individualised, have staff with seven key attributes (e.g., flexibility) and care organisations should fulfil 12 key attributes (e.g., timely provision). Also, in Australia, Hutchinson et al. (2020)  Knowledge about what works in practice is crucial to inform services. Appreciative inquiry (AI; Cooperrider et al., 1987), originally developed to create ideas for organisational change by focusing on what works well, is consistent with solution-focused research aims.
McCarthy (2017) suggested AI as potentially useful for resolving care issues usually labelled as demonstrating 'behavioural and psychological symptoms'. He illustrated the application of all four phases (discovery, dream, design, and destiny) of AI to find a helpful solution in a fictional case. Scerri et al. (2015) applied an adapted version of AI using only the discovery phase, to establish care processes that led to positive experiences for people with dementia on in-patient wards; then built on this, through staff workshops addressing the remaining phases, to gain improvements in care (Scerri et al., 2019).
Bearing in mind the lack of evidence on the nature of positive services, and the importance of hearing the voices of people with YOD (Brown et al., 2012;Johannessen & Möller, 2013;Mayrhofer et al., 2018), we aimed to contribute to the literature on how to provide effective YOD services by using an adapted AI approach to gather and analyse examples of helpful service support, from people with YOD and family carers.

| DE S I G N AND ME THODS
We used a qualitative design based on the discovery and dream phases of AI. The discovery phase involved gathering and analysing What is known about this topic?
• People with young onset dementia have distinctive needs.
• Most dementia services are 'all-age' and do not provide well for people with young onset.
• There is little UK research evidence on what sort of services are helpful.

What this paper adds?
• People with young onset dementia: data from people living with YOD and carers to understand what works well in support services; and the dream phase involved interpreting these findings to inform recommendations for good practice.

| Participants
Those with dementia needed to have received their diagnosis before the age of 65 and could have any dementia diagnosis, defined by DSM-5 (American Psychiatric Association, 2013), except for those dementia types usually served by other specialist services (see the introduction. Diagnosis was confirmed by services for those recruited via NHS sites and by self-reports for others. Carers could be any age and needed to self-identify as a 'family member/supporter' of a person with YOD, with details gathered about the amount of care provided (see  for details).

| Survey
We used a cross-sectional nationwide survey, to gather sociodemographic data, information on service use and satisfaction, and experiences of helpful support, between August 2017 and September 2018. The survey was developed by the research team.
It was piloted twice using cognitive interviews (Willis, 2004) with members of the Patient and Public Involvement (PPI) Group to refine wording, content, length and format. This article reports on the analysis of helpful support services, with other aspects reported elsewhere .
Participants were provided with briefing notes, followed by questions ( Table 1). The briefing notes were developed with the PPI group to help participants to recall their experiences and gave examples of general outcomes that might be helpful. Questions were presented in pairs, combining a closed question, to direct attention to a time-period, and an open question asking for a free-hand description of 'when the help and assistance you received was helpful'. The questions addressed three time periods: (a) between seeking help Section 1. Positive experiences of services and support We would like to know about good examples of help and assistance that have benefited younger people with dementia and/or family members/supporter(s). These could include: 1. A service or form of support which has: • helped you to maintain or improve your quality of life or, • made your life easier or, • addressed a concern you may have This could include, for example: a. help with your finances, b. your health or, c. maintaining activities that are important to you.
2. Advice or information. This could include, for example: • information about dementia, • about how to cope with specific symptoms or, • how to find the right services to meet your needs.
3. The approach or attitudes of a person who has provided help and assistance that has benefited you.
Survey sample question: During the year after diagnosis, can you think of times when the help and assistance you received was helpful?

Yes No
If you answered yes, please provide a detailed description of these experiences.

| Survey
Socio-demographic data were analysed using SPSS 23. Inductive thematic analysis (Braun & Clarke, 2013) was used to analyse the free-text responses. Emergent codes were clustered into three groups concerning what was provided, how it was provided and why it was helpful. This

| Interviews
Interviews were audio-recorded, transcribed verbatim, anonymised and analysed via template analysis (Brooks et al., 2015), using the survey themes as the initial template. Meaningful segments were labelled using the template and new emergent codes were added to existing themes or clustered into additional themes.
Interview analysis took place alongside data collection. Initially, three co-authors (VS, JLF and JO) independently read three transcripts and engaged in preliminary coding, leading to modification TA B L E 2 Interview guides used to elicit in-depth information on the nature of helpful services for young onset dementia Interview Version 1 (used with those at ease recalling survey information) Open discussion about the service examples provided in the survey. Where possible explore the needs of the person with young onset dementia and the family members/supporters separately. Then explore any additional examples of helpful experiences with services.
For each example given ask: Can you tell me about the events/ experiences you had leading up to being referred to/receiving this service? (prompts include) • What prompted the referral?
• What were you feeling/ needing at the time?
• Tell me about the process that you went through to get the service. • What happened when you saw them/ began to receive the service? • What specifically was good about it? • Why was it helpful to you? • How was the person's/services approach helpful? • How did it make you feel? • How was this different from how you felt before?
Version 2 (used with those who found it hard to recall survey information) I would like to understand what is important in the lives of people living with young onset dementia and their families and supporters since receiving the diagnosis of dementia. Some people find that life changes quite significantly and that some aspects of their lives become more important. Others describe how important it is to keep living life the way you have always done.
I would like to know what is important to you now in your life?
I would like to discuss what helps you to achieve the things in your life that are important to you. Some people find that they require help and assistance to live well with dementia and achieve the things that are important to them.
I would like to know what help and assistance you have received that has helped to you achieve the things that are important to you?
(i) What did they do for you? (ii) How have they helped you? (iii) What exactly was helpful about that? (iv) Why has it been helpful to you? of the template, which was then applied to successive sets of three interviews. Further refinements were made by consensus as necessary, following coding of each set. Interviews and analysis continued until data saturation was reached. The final template was then re-applied to the whole dataset by VS to check no further refinements were required. The inductively derived themes addressed the discovery phase of AI, which appreciates what is done well.

| Derivation of superordinate themes
In the dream phase of AI, we considered the themes from the perspective of their implications for YOD service provision. A researcher (VS) and two experienced clinical-academic researchers (JLF and JO) interpreted the implications of each theme for service provision and a set of superordinate themes was agreed through discussion and consensus.

| Quality assurance
Specific quality criteria for qualitative research are contested (Bryman et al., 2008). However, to enhance confirmability, we paid attention to reflexivity through keeping detailed reflexive notes and holding regular team discussions and coding seminars. We established credibility by discussing the themes with the PPI group, wider research team, Steering Committee and external audiences to obtain feedback on plausibility.

| RE SULTS
Of 233  ples, which did not reveal new meanings or themes (Bowen, 2008).
Twenty-four follow-up interviews took place: eight with people with YOD, two with dyads and nine with carers using the open approach, and five with carers using the semistructured approach (see Table 3 for their socio-demographic data). As two interviews were with dyads there were 26 participants in total.
Analysis led to 12 themes, grouped into three superordinate themes reflecting the level at which they could be best enacted in practice: 'Organisational coherence', 'Functional consistency' and 'Person-centredness' (see Table 4). Figure 1 gives a diagrammatic representation, with the superordinate themes presented as reflecting macro, meso and micro levels of practice. Each superordinate theme and its associated themes is described below.

| Organisational coherence
This superordinate theme incorporates three themes with implications for overall coherence of YOD services: the helpfulness of integrated services, of care being managed in specialist services and of service consistency over time.

| Functional consistency
This super-ordinate theme groups together four themes (ageappropriate, holistic, responsive and accessible) related to the way helpful services were organised and delivered to meet the needs of those affected by YOD.
Participants emphasised the significance of services that pro- It is apparent that this service was valued due to its provision for different aspects of the condition, as well as addressing the needs of all family members, including children.
Participants also emphasised the 'responsiveness' of helpful services, through providing timely responses as needs changed: TA B L E 4 The nature of helpful services for young onset dementia: super-ordinate themes, main themes and theme descriptions derived from a survey (n = 233) and interviews (n = 24)

Super-ordinate themes Themes Brief theme description
Organisational coherence Integrated care Services provided collaboratively, rather than in silos, organised in a way that enables a coherent tailored care pathway Care managed by specialists Care provided and co-ordinated by specialist services and professionals who are trained and experienced with YOD

Consistent
Services that give the person regular and continuing contact with a known person and/or service, who remains in touch as needs change over time

| Person-centredness
The third superordinate theme contains five themes (positive attitude, collaboration, flexibility, in-person support and user-friendly material) that captured aspects of person-centred interaction with care providers.
Professionals were seen as being helpful when they conveyed a respectful, compassionate, attentive and genuine 'attitude'. These positive attitudes appeared to lay the foundation for 'therapeutic collaboration', expressed here by a carer: The social worker assigned to us began to think creatively about the care needed rather than following the normal path associated with an Alzheimer's di-

| D ISCUSS I ON
In this study, we used an adapted version of AI. Through the discovery phase, we inductively developed twelve themes about the nature of helpful YOD services and, through the dream phase, we interpretively grouped these into superordinate themes at micro, meso and macro levels, to draw out practice implications. These levels have been widely applied to conceptualise facilitators and barriers to successful service delivery, including in dementia care (Meiland et al., 2004;van Mierlo et al., 2014).
The micro level relates to the person-centred nature of helpful interaction between professionals and people with YOD/carers. The helpful facets of positive attitude that we identified correspond closely to those that have been found important to people of all ages with dementia, resonating with the theory of person-centred dementia care (Brooker, 2019) the human-rights approach (Cahill, 2018) and the dimensions of a core outcome set for assessing the quality of psychosocial community-based interventions (Reilly et al., 2020). It is therefore not surprising that these qualities emerged as central in our study or that they have been found within other YOD research (Johannessen & Möller, 2013;Westera et al., 2016). Possibly more particular to the YOD population was the emphasis on flexibility, also identified in Australian studies (Brown et al., 2012;Cations et al., 2017;Westera et al., 2016). This may be especially important to those under 65 years, who usually have many ongoing responsibilities, varied needs and ambivalence about receiving support, due to stigma attached to untimely experience of dementia. Collaborative partnerships between those with YOD/carers and professionals, also identified as important by Mayrhofer et al. (2018) and Johannessen and Möller (2013), may equally be particularly valued by those in middle-age, who have not yet reached an age when they are reconciled to relinquishing control to younger generations. Support delivered via a collaborative, continuing relationship has been found in previous research to engender trust and prevent isolation (Roach & Drummond, 2014) and enables services to be responsive to changing needs (Carone et al., 2016). The value placed on in-person support was unexpected and has become particularly noteworthy at this time of the Covid-19 pandemic (British Psychological Society, 2020;  when social distancing has led to the suspension of social groups and activities. At meso level, we grouped aspects of support that reflected functional aspects of services. The need for age-appropriate, holistic support has been stressed over many years in previous YOD research (Beattie et al., 2004;Hutchinson et al., 2020;Mayrhofer et al., 2018), and the needs we identified for services to be responsive and accessible have also been previously reported (Bentham & La Fontaine, 2005;Brown et al., 2012;Cations et al., 2017;Mayrhofer et al., 2018;Westera & Fildes, 2014). However, only two studies were UK-based of which one was localised (Bentham & La Fontaine, 2005), and the other had relatively small samples (Mayrhofer et al., 2018). Our study therefore amplifies existing knowledge by virtue of the large set of examples provided directly by people with YOD and carers.
At macro level, our themes are consistent with valued organisational attributes in the Australian context (Westera & Fildes, 2014).
Other findings of the Angela project highlight that specialist services are not only subjectively valued but also outperform other service delivery models on quality indicators . Consistency of professional and team input was very highly valued by participants, who usually require support over a number of years. Yet this can be interrupted by service reorganisations due to short term funding or commissioning (Mayrhofer et al., 2018) or be prevented when only short-term follow-up is commissioned from memory services (Giebel, Eastham, et al., 2020).
Although this factor could have been placed at meso level, we included it at macro level as, without long-term commissioning, it is not possible to have the stability needed for service consistency over time.

| Limitations
Our survey carried a risk of selection bias, towards educated participants in contact with services. Despite targeting recruitment towards Black, Asian and Minority Ethnic (BAME) communities via social media and specialist third-sector organisations, we were only able to recruit a few participants from BAME communities. The briefing notes, used in the survey, may have primed recall of particular features of services.
However, the notes were general, and the majority of participants' responses were specific and detailed. Additionally, no one under 50 years when diagnosed was interviewed, although YOD prevalence is higher in the 50s than at earlier ages (Vieira et al., 2013).

| Implications
The YOD-related essence of our themes may become diluted when raised to high level labels. However, the first-hand words and ex-  , 2016). This has been recognised in the Netherlands (Carter et al., 2018) and should be emulated in the United Kingdom.

| CON CLUS IONS
To the best of our knowledge, this is the first large-scale UK study to report on specific features of helpful services from the perspective of people with YOD and carers. Over the past 5 years, research has moved from identifying needs in YOD to considering what needs to change to meet these needs. Our findings consolidate and expand on this work. The AI approach allowed us to take inductively-derived themes and use them to envision the implications for how they could be enacted, leading to a three-tiered set of recommendations for good practice at the levels of service design, function and delivery.

ACK N OWLED G EM ENTS
The researchers acknowledge the support of the National Institute

CO N FLI C T O F I NTE R E S T
We have no conflict of interest to declare.

AUTH O R CO NTR I B UTI O N S
JO was involved in study design, analysis, drafting and re-drafting of the article. VS was involved in study design, data collection, analysis, drafting and re-drafting of the article. JLF was involved in study design, data collection and analysis, and commenting on article drafts.
MOM, BJ, JP and JC were involved in study design and commenting on article drafts.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.